Along-Axis Segmentation and Growth History of the Rome Trough in the Central Appalachian

25 Pages·2000·1.43 MB·English

Along-Axis Segmentation and Growth History of the
Rome Trough in the Central Appalachian Basin1
Dengliang Gao,2 Robert C. Shumaker,3 and Thomas H. Wilson3
ABSTRACT Computer-aided interpretation of seismic data
and subsurface geologic mapping indicate that the
The Rome trough, a northeast-trending graben, is Rome trough experienced several major phases of
that part of the Cambrian interior rift system that deformation throughout the Paleozoic. From the
extends into the central Appalachian foreland basin Early(?)–Middle Cambrian (pre-Copper Ridge deposi-
in eastern North America. On the basis of changes tion), rapid extension and rifting occurred in associ-
in graben polarity and rock thickness shown from ation with the opening of the Iapetus-Theic Ocean
exploration and production wells, seismic lines, and at the continental margin. The Late Cambrian–
gravity and magnetic intensity maps, we divide the Middle Ordovician phase (Copper Ridge to Black
trough into the eastern Kentucky, southern West River deposition) was dominated by slow differen-
Virginia, and northern West Virginia segments. In tial subsidence, forming a successor sag basin that
eastern Kentucky, the master synthetic fault zone may have been caused by postrift thermal contrac-
consists of several major faults on the northwestern tion on the passive continental margin. Faults of
side of the trough where the most significant thick- the Rome trough were less active from the Late
ness and facies changes occur. In southern West Ordovician–Pennsylvanian (post-Trenton deposi-
Virginia, however, a single master synthetic fault, tion), but low-relief inversion structures began to
called the East-Margin fault, is located on the south- form as the Appalachian foreland started to devel-
eastern side of the trough. Syndepositional motion op. These three major phases of deformation are
along that fault controlled the concentrated deposi- speculated to be responsible for the vertical stack-
tion of both the rift and postrift sequences. The ing of different structural styles and depositional
East-Margin fault continues northward into the sequences that may have affected potential reser-
northern West Virginia segment, apparently with voir facies, trapping geometry, and hydrocarbon
less stratigraphic effect on postrift sequences, and a accumulation.
second major normal fault, the Interior fault, devel-
oped in the northern West Virginia segment. These
three rift segments are separated by two basement INTRODUCTION
structures interpreted as two accommodation zones
extending approximately along the 38th parallel The Rome trough (Woodward, 1961; McGuire and
and Burning-Mann lineaments. Howell, 1963) is one of the major rift elements of the
interior rift system (Harris, 1978) that formed in east-
ern North America during the Early and Middle
©Copyright 2000. The American Association of Petroleum Geologists. All
Cambrian in association with the opening and
rights reserved.
1Manuscript received March 9, 1998; revised manuscript received April spreading of the Iapetus-Theic Ocean (Thomas,
13, 1999; final acceptance June 28, 1999. 1991) (Figure 1). The Rome trough extends across
2Marathon Oil Company, P.O. Box 3128, Computer-Aided Interpretation,
extensive areas of oil and gas production in the
Houston, Texas 77056; e-mail: [email protected]
3Department of Geology and Geography, West Virginia University, central Appalachian foreland basin. The Rome
Morgantown, West Virginia 26506. trough has aroused interest among structural geolo-
This study was supported by the National Research Center for Coal and
Energy (NRCCE). We thank Columbia Natural Resources (CNR) Inc. for gists and petroleum geologists over the past 30 yr
allowing the publication of the seismic lines. Acknowledgment is made to the (e.g., Woodward, 1961; McGuire and Howell, 1963;
West Virginia Geological and Economic Survey for well data and the sonic
Harris, 1975, 1978; Ammerman and Keller, 1979;
and density logs for subsurface mapping and synthetic seismic modeling. We
also benefited from the application of the computer software packages of the Kulander and Dean, 1980, 1993; Shumaker, 1986a, b,
MCS (Mapping-Contouring System), SURFACE III, DEAM (Data Editing and 1993, 1996; Kulander et al., 1987; Thomas, 1991; Patchen
Management), DIG (Digitization), and the programs for synthetic seismic
et al., 1993; Drahovzal, 1994; Wilson et al., 1994a, b; Gao,
analysis. This paper benefited from suggestions by Jeffrey A. Karson (Duke
University) and journal reviews by William A. Thomas, an anonymous 1994; Gao and Shumaker, 1996; Harris and Drahovzal,
reviewer, and the elected editor Neil F. Hurley. Thanks also go to Sharon L. 1996; Shumaker and Wilson, 1996; Ryder et al., 1997a,
Crawford, Thomas R. Evans, and Charles A. Meeder (Marathon Oil
Company) for their support. b; Beardsley, 1997). Although thestratigraphy and
AAPG Bulletin, V. 84, No. 1 (January 2000), P. 75–99. 75
76 Rome Trough
Figure 1—Tectonic map
showing the Iapetian
structure of southeastern
North America and the
Appalachian foreland
basin (after Shumaker and
Wilson, 1996; Shumaker,
1996). Rome trough and
other grabens of the
interior rift system are
from Shumaker (1986a).
Rifts and transform faults
at the plate margin are
from Thomas (1991).
The 38th parallel
lineament is based on
Heyl (1972), and the
Burning-Mann lineament
is based on Shumaker
(1986b). Also shown are
the locations of areas
shown in Figures 3 and 4.
structure of the Rome trough have been extensive- collectively part of a more extensive interior rift
ly discussed in previous studies, little has been pub- system (e.g., Shumaker, 1986a, 1996; Shumaker
lished on the along-axis segmentation in trough and Wilson, 1996). This system formed in associa-
geometry and its control on focused sedimentation tion with the late-stage opening of the Iapetus-
and hydrocarbon accumulation. In this study, we Theic Ocean at the plate margin during the Early
compare and contrast the subsurface geology of and Middle Cambrian (Rankin, 1976; Thomas,
different segments on the basis of well, seismic, 1977, 1991; Read, 1989; Ryder et al., 1997a, b).
magnetic, and gravity data. The results reveal com- The crustal extension of the interior rift system
plexities in the geometry and growth history of produced a thick sequence of lower Paleozoic sedi-
the trough that have enhanced our understanding mentary rocks in several grabens (Shumaker,
of the mechanism for hydrocarbon entrapment 1986a). The Rome trough (McGuire and Howell,
and tectonic evolution of the pre-Appalachian and 1963; Harris, 1975) is one of the elements of the
Appalachian foreland. That understanding, in turn, interior rift system that extends into the
should provide the basis for a more accurate Appalachian foreland in eastern Kentucky and
assessment of the hydrocarbon potential of both western West Virginia where it follows the north-
deeply buried and shallow Paleozoic reservoirs east-trending magnetic gradient named the New
along the trough. York-Alabama lineament (King and Zietz, 1978).
Thousands of meters of sedimentary rocks are pres-
ent within the trough, which consists of diverse
GEOLOGIC SETTING lithologic units (Figure 2) (Schwietering and
Roberts, 1988; Ryder, 1992). These sedimentary
The central Appalachian foreland basin is under- rocks can be divided into rift (Rome-Conasauga),
lain by a series of continental grabens that are passive-margin (Copper Ridge-Black River), and
Gao et al. 77
south to north, the Rome trough is interrupted by a
M west-trending basement fault zone, named the 38th
TE SERIES GROUPS OR FORMATIONS,
YS MEMBERS AND BEDS parallel lineament, that has been observed at the
S surface (Heyl, 1972) and by a north-trending mag-
netic gradient, named the Burning-Mann lineament,
UPPER MAUCH CHUNK GROUP
N that extends from the Burning Springs to the Mann
A
PI Mountain anticline developed at the surface
SIP MIDDLE GREENBRIER LIMESTONE (Shumaker, 1986b). These two basement linea-
S
SI MACCRADY FORMATION ments (Figure 3) divide the Rome trough into the
MIS LOWER POCONO GROUP BiWg eInirjun eastern Kentucky, southern West Virginia, and
Berea northern West Virginia segments.
Sunbury Sh.
AN UPPER OHIOSHALECLBEEVDEFLHOAUCRNRHDDOA SNGMH RMEAIMNELM BESEHBRAERLE DU EN UVDPOIVPNEIIDRAEND ACADIAN FORELAND sVgteirrsSgutheicnudtim uat hraaeakntsed mr ot ahfon estdithr o eWaf f Aiftlehspcoeptn loaa (lnr1ag 9csee9hrd 6ibia)ma ndse eifsnmoctruaeetsnilsotae nfnda.d ubT lahitnsse e iyWnm s teeuhnsgett-
NI JAVA FORMATION
O WEST FALLS ANGOLA SHALE MEMBER Rome trough formed during the Precambrian
V
E FORMATION RHINESTREET SHALE MEMBER Grenville orogeny. Subsequently, the trough experi-
D
MIDDLE ONONDAGA HUNTERSVILLE enced rifting during the Early(?) and Middle Cam-
OUS LIMESTONE CHERT brian, postrift subsidence (Ryder et al., 1997a), pos-
LOWER R ORISKANY SANDSTONE
FE HELDERBERG GROUP sibly forming a sag basin in the passive-margin
UPPER NI SALINA FORMATION sequence (Figure 2) during the Late Cambrian and
R
URIAN MIDDLE CO NLKEOWECEKBFUPEORRRG ST AS DANONDLDSOTSMOTONITNEEE Oasrsdoocviaictieadn ,w aintdh, af ifnoarlelyl,a bnrdo bada srieng siotangael sduubrsiindge nthcee
L C
SI ROSE HILL FORMATION NI middle and late Paleozoic.
LOWER TUSCARORA SANDSTONE CO Basement structures of the Rome trough in eastern
JUNIATA FORMATION A
VICIAN UPPER MWBALETRALRTLCEISKNN C STRRBOIVEUNEER KLRGI MFFFOOOERRRSMMTMOAAANTTTEIIIOOONNN T KBanleadnc ktBu eectak raydl .s,w l1ee9yr,7e 16 d;9 iA8sm4c;um Bsselearcdmk ia,n n1 pa9nr8ed6v ;Ki oDeulrlase hrs,o t1uv9dz7aie9l,s; 1C(9ea9.bg4l.e,;
O MIDDLE EN
ORD LUOPWPEERR KNOXGROUP BCEOERSPKTOPM.ES PAERE N RTRTUEIDORNGW SSEANA FN NFODDORSSRMTTMOAOANTNTEIEOIONN PASSIVMARGI Deaxrreaaamhl opevlxezt,ae Aln mat nmadne dNrm ogaegnoe rma, n1ed9t rK9y5e lo;l efS rht (hu1em9 7Ra9ko)em dr,ei s1 ct9ur9oss6ue)gd.h Ft hoiner
N
RIA MIDDLE CONASAUGA FORMATION eastern Kentucky using gravity and deep drilling
AMB TORMOSMTOE WFONR DMOALTOIOMNITE RIFT dbaatsae.m Tehneti rf agurlatvsi tcyo mntorodlelleidn gt hree sgurlatsb einnd giceaotme ethtrayt
C LOWER
? BASAL SANDSTONE and thickness of the Paleozoic sedimentary rocks in
N C
AMBRIA PPEREROZOI IOGFN TEHOEU BS AASNEDM MENETTAMORPHIC ROCKS t(mh1ai9sj9 op5ra) s rautn bodsf uStrhhfeua cmReoa fkmaeuerl tt(sr1o 9iun9g 6eh)a. s mDtearrpanhp oeKvdez ntahtl ueac nekdxy t Neunostig neogrf
C UT
PRE PRO eoxf iwstiidnegs gperoealodg riecg mioanpasl, sdeeisemp iwc elilnl edsa.ta, and a series
The fundamental geological work in West
Figure 2—Generalized stratigraphic sequences of the
Virginia dates back to the early 1900s, when the
study area (after Schwietering and Roberts, 1988; Shu-
West Virginia Geological and Economic Survey
maker and Wilson, 1996).
published a series of geologic reports and maps.
This was followed by extensive discussions on the
structural geology and stratigraphy in West Virginia
foreland (above Trenton) sequences (Shumaker (e.g., Neal and Price, 1986; Caramanica, 1988;
and Wilson, 1996). Schwietering and Roberts, 1988; Zheng, 1990;
Gravity maps (Ammerman and Keller, 1979; Shumaker, 1993; Shumaker and Coolen, 1993;
Kulander et al., 1987) and a total magnetic intensity Wilson et al., 1994a, b; Gao, 1994; Gao and Shu-
map (King and Zietz, 1978) indicate that the Rome maker, 1994, 1996). For example, Shumaker (1993)
trough shows significant along-axis variations in and Shumaker and Coolen (1993) reported on the
depth of the rift valley and thickness of overlying studies on the East-Margin fault and the 38th paral-
Paleozoic sediment infill. Superposition of the total lel lineament. Wilson et al. (1994a, b) reported on
magnetic intensity map with major basement struc- the study of reflection seismic data across the
tures of the Rome trough (Figure 3) indicates the Rome trough of northern West Virginia. Their seis-
spatial relationship between magnetic intensity mic analyses indicated that basement faults and
variations and major basement lineaments. From fault-bounded blocks have been intermittently active
78 Rome Trough
Figure 3—Superimposed magnetic and tectonic map showing the spatial relationship among the total magnetic
intensity, areal extent of the three major segments of the Rome trough, and associated major basement fault sys-
tems. See Figure 1 for location. Magnetic intensity map is from King and Zietz (1978). Basement faults in eastern
Kentucky are from Ammerman and Keller (1979). Basement faults in northern West Virginia are based on Shumak-
er and Wilson (1996). Note that the major magnetic intensity lows are spatially related to the three major segments
of the trough, which are named the eastern Kentucky segment, the southern West Virginia segment, and the north-
ern West Virginia segment.
during the Paleozoic and have significantly affected oblique (wedge-shape) transfer fault system that
the deposition of the Paleozoic sedimentary rocks in accommodated the complex deformation of the
the Rome trough of northern West Virginia. subsurface structures in southwestern West Virginia.
Using more than 4000 shallow wells and several In this study, using new seismic data and inter-
seismic lines, Gao (1994) and Gao and Shumaker pretation and subsurface mapping techniques, we
(1996) mapped the subsurface geology in south- compare and contrast structures among the three
western West Virginia and documented the geomet- rift segments to evaluate the along-axis variation in
ric and kinematic relationship of shallow structures graben geometry. We propose that the 38th parallel
to the East-Margin fault and the 38th parallel and and Burning-Mann lineaments are two possible
Burning-Mann lineaments. Based on their spatial and accommodation zones to transfer extension from
temporal relationship, Gao (1994) and Gao and one graben segment to the next along the trough.
Shumaker (1996) suggested that the 38th parallel We establish a structural model to emphasize along-
and Burning-Mann lineaments represent a possible rift segmentation and geometric and kinematic
Gao et al. 79
82°W Figure 4—Index map
showing the data set in the
N southern West Virginia
segment of the Rome
Line 9 trough. See Figure 1 for
-4100 location. The approximate
locations of seismic lines
Line 6
are indicated by bold straight
-4000 0-4400 lines. The background
Southern WV Segment 440 structure contour map,
- which was constructed
on the basis of more than
-3900 1000 shallow wells, shows
-3800 the top of the Devonian
-3700
L1469 -3600 K3462 Line 5 Osonuothnedrang aW Leimst eVsitrognine iian
(from Gao and Shumaker,
LincolnL Cinoeun 8tyLi-n31e0 -703300-34W0ar0fi-3el5d 0a0nticline East-Margin Fault -47Kana0wha 0County 1afisanotsn9rr t9aaue 6el rcyrx)ptes.ut ofirTrselaeh.arpsetei on vnOlcigaanet d iotsnrenuegerdn pfaada ngc-dsaeu srefravcees
-3000 Boone County
Line 3 M805
Line 4
Line 10
Line2
38°N 38°N
Line 11 -3900 Logan County Line 1 C.I. = 100 ft
0 25000 50000 ft
Mingo County 0 5 10 15 20 25 km
Line 12
82°W
differences among the three rift segments, which Sonic and density logs of three deep wells (Figure 4)
may have important implications for hydrocarbon were digitized. Impedance, reflection coefficients,
exploration along the Rome trough and other rift and normal incidence synthetic seismograms were
systems on a regional basis. computed (Figure 5). The synthetic seismograms
are derived from the convolution between reflec-
tion coefficients and a zero-phase wavelet. The
SOUTHERN WEST VIRGINIA SEGMENT stratigraphic positions of the major reflection
events are shown in Figure 5. We identified strati-
The southern West Virginia segment is defined graphic intervals based on drillers’ logs and lateral
between the 38th parallel and Burning-Mann linea- correlation with other log data (e.g., Ryder, 1992;
ments (Figure 3). Because the southern West Drahovzal and Noger, 1995), and the seismic event
Virginia segment is bounded by the 38th parallel picks are based on synthetic seismic correlation
and Burning-Mann lineaments to the south and with actual seismic lines (e.g., Figure 5c).
north, respectively, a detailed study of the geometry To better analyze and emphasize lateral and tem-
and growth history of that segment, using new seis- poral variations in structural geometry, we digi-
mic data and interpretation techniques, provides a tized two-way traveltimes of major seismic events
key to understand the along-axis variation in graben associated with several key horizons. The digitized
geometry and growth history of the Rome trough. reflections were vertically exaggerated by upward
shifting of sequential horizons coupled with rescal-
ing on the vertical axis (Wilson et al., 1994a, b;
Seismic Analysis Shumaker and Wilson, 1996). Although vertical
shifts of reflection events eliminate absolute values
A total of 12 seismic lines were interpreted in of arrival time, such shifts do not affect absolute
the southern West Virginia segment (Figure 4). differences in structural relief along individual
80 Rome Trough
2 inandom-ajorsec-
Impedance (10)Trace4 4.08.00.00.0Actual DataImpedance (10)Trace4Impedance (10)Trace4(from line 6)Top Trenton4.00.08.00.0Top Black River0.04.08.00.00.0 Top Beekmantown1.0Top Wells CreekTop Rose RunTop Copper RidgeTop Beekmantown 0.25Top Rose RunTop Big LimeTop Conasauga1.43Top Copper RidgeTop Berea1.3Top Conasauga 0.501.68 1.6 )))sss((( 0.75Top Onondaga1.93Top OriskanyTop RomeTop Newburg1.9Top Rose HillTop TuscaroraTop Rome1.0 2.18 2.2Top TrentonTop TomstownTop Black River1.25 Top TomstownTop Basement2.43Top Wells Creek(a)(c)Top BeekmantownTop Basement2.5(b) Figure 5—Impedance and synthetic seismograms for (a) well L1469 in Lincoln County, (b) well M805 in Mingo County, and (c) well K346Kanawha County. Wells show their lateral correlation and similarity in interval velocity. Impedance is calculated by merging the sonic (velocity) density logs. Note the similarity in velocity package within the same stratigraphic interval. The normal-incidence synthetic seismograms were cputed from the convolution between the reflection coefficient (calculated from acoustic impedance profile) and zero-phase wavelet. Several mlithologic boundaries can be resolved as separate reflections. Example correlation between a synthetic seismogram and an actual seismic time tion is shown in (c). Time is in seconds two-way traveltime.
Gao et al. 81
Ao Ao
TimOeri gSiencatlion 0.5 Bo TimOeri gSiencatlion 0.5 Fault (?) BCoo
Anticline (?) Co
A
B
TimSeh iSfteecdtion 0.1 AB TimSeh iSfteecdtion 0.1
Fault
Anticline C
C
Time B-A Time
Difference 0.1 C-B Difference 0.1 B-A
C-B
(a) (b)
Figure 6—Two simplified examples showing the benefits of constructing shifted arrival times and time difference
curves from the original time section. Subtle structural and stratigraphic features such as the (a) anticline and the
(b) fault are difficult to recognize in the original time section. By shifting sequential reflections and increasing the
vertical scale, the geometry of the anticline (a) and the location of the fault (b) are easily recognizable; furthermore,
the time difference curves, calculated by sequentially subtracting shifted arrival times, remove the structural effect
to emphasize the differential subsidence associated with the drape anticline (a) and the growth fault (b).
reflectors. Figure 6 demonstrates how vertical and the location of depocenters. Unlike arrival
exaggeration is achieved via vertical shifts and time, lateral variation in time difference is not
rescaling of the sequential horizons. Original, affected by lateral variations in velocity of the over-
unshifted reflections generally appear as widely lying strata. Basically, lateral variations in time dif-
spaced flat areas. The original plotting scale tends ference represent changes in rock thickness in the
to mask or attenuate subtle structural and strati- absence of significant lateral variation in interval
graphic features. By upward shifting and rescaling velocity. After careful examination of sonic logs of
on the vertical axis, subtle variations associated several wells, and after making several test calcula-
with the anticline (Figure 6a) and the fault (Figure tions, we note that average interval velocity is rela-
6b) that would otherwise be difficult to recognize tively constant based on sonic logs from the three
are exaggerated. In the study area, we found no deep wells; therefore, variations in time difference
direct evidence that supports the presence of veloc- can be properly interpreted as related to thickness
ity anomalies, and both seismic lines and sonic logs changes, and local depocenters can be better
indicate that major thickness changes within indi- defined with the help of the differential subsi-
vidual lines are restricted to the deepest, synrift dence curves.
reflections. Most of the shallow intervals show rela- Line 1 is located in the central part of the south-
tively minor changes in thickness; hence, in the ern West Virginia segment, away from both the
absence of a satisfactory time-depth conversion 38th parallel and Burning-Mann lineaments. The
table, major lateral variations in arrival time can be digitized and processed section (Figure 7a) sug-
properly interpreted as related to structural relief. gests that this segment of the Rome trough is an
The distortion, if any, may occur at the basement asymmetric graben with a single East-Margin fault
structure level due to significant thickness variation dipping toward the northwest on the southeastern
of the overlying rift sequence. side of the trough. Reflections, interpreted as tops
To remove structural effects and to emphasize of the Rome, Tomstown, and basement units, dip
differential subsidence, we constructed a series of toward the East-Margin fault, suggesting that the
subsidence curves by calculating the differences in hanging wall rotated clockwise into the footwall.
arrival times between the adjacent horizons, called We found more than 0.5 s of lateral variation in
time difference (Figure 6), that aid in identifying time difference (Figure 7b) across the East-Margin
changes in the polarity of differential subsidence fault in the rift sequence (see Rome-Conasauga
82 Rome Trough
Figure 7—(a) Digitized and LINE 1
enhanced arrival times
and (b) time difference NW SE
curves of line 1 (see
Top Berea
Figure 4 for location and
scale) showing lateral TTroepn Ttornenton OTonpo nOdnaognadaga
variations in structure and
Conasauga
differential subsidence, Top Conasauga
T
respectively. Note no W
absolute values are T) Top Rome
attached to the vertical e (s
axis because of vertical ativ1.0 Top Tomstown East-Margin Fault
sahnidf tt iomf eth dei fafrerrievnacl et icmuersves. Rel Sequences
See text for explanation. Acadian Foreland
TWT = two-way traveltime. Taconic Foreland
Top Basement
Passive Margin
Rift
(a)
NW SE
Onondaga-Berea
e
c
n Trenton-Onondaga
e
er Conasauga-Trenton Rome-Conasauga
f
f
Di ) Tomstown-Rome
e (s
m 0 Basement-Tomstown
Ti 1.
ve East-Margin Fault
ati
el
R
(b)
time difference), but only 0.2 s difference in the Near the 38th parallel lineament is a series of
passive-margin sequence (see Conasauga-Trenton east-west–trending basement faults. At the border
time difference) and 0.1 s difference in the fore- between the southern West Virginia and eastern
land sequence (see Trenton-Onondaga time differ- Kentucky segments, lines 2, 3, and 4 show a well-
ence), indicating that a major amount of differen- defined basement fault that juxtaposes basement
tial subsidence and deposition across the East- rocks on the south with sedimentary rocks of the
Margin fault occurred during the Early(?) and trough on the north (Figure 8). The changes in
Middle Cambrian. The thickest part of the Conasauga- both dip polarity (Lee, 1980) and strike of this base-
Onondaga interval is located on the hanging wall ment fault (Gao, 1994; Gao and Shumaker, 1996)
just northwest of the East-Margin fault, but unlike indicate the structural complexity at the border of
the underlying rift sequence, deposition is not the two rift segments.
restricted within the trough but extends outside Near the Burning-Mann lineament, internal struc-
of the trough. The overlying Onondaga-Berea tures of the trough are more complicated. Digitized
interval reveals a shift in the depocenter toward reflections from line 5 (Figure 9a) reveal that reflec-
the southeast (Figure 7), which suggests inver- tors above the Trenton Limestone dip to the east,
sion across the East-Margin fault following depo- whereas reflectors below the Trenton Limestone
sition of the Devonian Onondaga Limestone. generally dip to the west. The East-Margin fault
The relative time difference plot (Figure 7b) appears to be a high-angle normal fault dipping to
indicates that normal displacement of the East- the northwest with a large normal offset of units in
Margin fault occurred intermittently throughout the rift sequence below the top of the Conasauga
the Paleozoic. Formation. The southeast-dipping reflections of the
Gao et al. 83
LINE 4 Figure 8—Seismic line 4
(see Figure 4 for location
and scale) across the
SE NW
East-Margin fault in the
southern West Virginia
Top Berea segment of the Rome
trough. Note the
0.5 East-Margin fault and
the discontinuity of
Top Onondaga
reflections across the
fault. TWT = two-way
Top Trenton traveltime.
1.0
Top Conasauga
)
s
T ( 1.5 Top Basement
W
T
2.0
Top Basement
East-Margin Fault
2.5
Rome Formation and the Tomstown Dolomite near Farther to the north, line 6 is located at the bor-
the East-Margin fault indicate a clockwise rotation der between the southern West Virginia and north-
of the fault block toward the East-Margin fault dur- ern West Virginia segments. Here, the trough is
ing the deposition of the rift sequence. Variations complicated by a major basement fault in the interi-
in the time differences (Figure 9b) from the base- or of the trough, called the Interior fault (Figure 10).
ment to Tomstown, the Tomstown to Rome, and The extension, which occurs largely across the East-
the Rome to Conasauga indicate southeastward Margin fault to the south, probably is distributed
thickening locally in the vicinity of the East-Margin between the two major faults toward the northern
fault. The Conasauga to Trenton interval shows a West Virginia segment north of the Burning-Mann
northwestward thickening into the Rome trough, lineament. The Interior fault shows a normal offset
which would be expected for a sag basin. The of more than 0.35 s, which is larger than that of the
Trenton to Rose Hill and the Rose Hill to Onondaga East-Margin fault (compare the Interior fault and
intervals both show slight thickening toward the the East-Margin fault in Figure 10a). The variations
northwest. These observations suggest that the in interval time difference (Figure 10b) suggest a
clockwise rotation of the basement block occurred complex history of differential subsidence during
mainly during the Early(?) and Middle Cambrian. the Paleozoic controlled by the Interior fault and
The Onondaga to Huron interval, however, thick- the East-Margin fault. Interestingly, the Interior fault
ens toward the southeast across the East-Margin seems to have a longer history of deformation than
fault, indicating a shift in subsidence polarity the East-Margin fault; the Interior fault influenced
toward the southeast after deposition of the the deposition from the Cambrian to Devonian
Onondaga. This shift is accompanied by a change with alternating polarity of differential subsidence,
from carbonate to clastic sedimentary rocks of the whereas the East-Margin fault had minor effect on
Acadian sequence that is generally considered to sedimentary rocks deposited after the Conasauga
mark the onset of an orogen in the growth history was deposited. These observations indicate that
of the Appalachian foreland. line 6 shows different internal structure and
84 Rome Trough
Figure 9—(a) Digitized LINE 5
and enhanced arrival
times and (b) time NW SE
difference curves of line 5 Top Huron
(see Figure 4 for location Top Trenton Top Rose Hill Top Onondaga
and scale) showing lateral
variations in structure and T Top Conasauga
W
differential subsidence,
raebsspoelucttiev vealylu. eNso ater eno ve T0 (s) Top Rome East-Margin Fault
attached to the vertical axis ati1. Top Tomstown Sequences
el
because of vertical shift of R Acadian Foreland
the arrival times and time Taconic Foreland
difference curves. See text Passive Margin
for explanation. TWT = Top Basement
Rift
two-way traveltime. (a)
NW SE
ce Onondaga-Huron
n Rose Hill-Onondaga
e
r Trenton-Rose Hill
e Rome-Conasauga
Diffs) Conasauga-Trenton Tomstown-Rome
e 0 (
m1. Basement-Tomstown
Ti
e
ativ East-Margin Fault
Rel (b)
growth history of the trough than line 5, which deposition decreases from south to north relative
suggests that graben geometry and growth history to the Interior fault. (2) Three vertically stacked
may have changed from the southern West Virginia structural styles are associated with three vertically
segment to the northern West Virginia segment. stacked rock sequences (Figure 2) that reflect
On the northwestern side of the Rome trough changes in the structural history of the study area.
(Figure 11), the top of basement dips gently east- These styles include a half graben associated with a
ward toward the East-Margin fault; the dip may be rift sequence that abruptly appears at the eastern
caused by the rotational subsidence at the western margin of the rift, a relatively broad sag basin asso-
margin of the trough. Increasing dip of the reflec- ciated with a passive-margin sequence that slightly
tions with depth indicates that rotational subsidence thickens above the rift, and a foreland sequence
largely occurred during the rift stage, and that differ- that expands eastward toward the plate margin and
ential subsidence or clockwise rotation became less is particularly apparent on the seismic expression
active after the formation of the rift basin. of the Acadian clastic wedge.
In summary, several points regarding the south-
ern West Virginia segment can be drawn from seis-
mic analysis. (1) The strike, dip, and throw of the Structure and Isopach Contouring
East-Margin fault change significantly along strike
from south to north. Between the 38th parallel and A total of 2221 control points were used to con-
Burning-Mann lineaments, the East-Margin fault struct subsurface geologic maps. These data
strikes to the northeast and dips to the northwest. include depths to the top of the Devonian
Toward the south at the segment border, the East- Onondaga Limestone extracted from the wells,
Margin fault swings abruptly to the west near the depths to the top of the Precambrian basement dig-
38th parallel lineament. Toward the north at the itized from an acoustic basement map (J. Lemon,
segment border of the Burning-Mann lineament, 1993, personal communication), and depths to sev-
the graben geometry is complicated by an Interior eral intervening horizons of the Ordovician and
fault, and the influence of the East-Margin fault on Cambrian converted from seismic sections.